From nobody Tue Oct 7 15:01:31 2025 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 294F62957CE; Wed, 9 Jul 2025 11:53:25 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1752062006; cv=none; b=abZ84yaI2KOAPMgDgXvzxuwDZK6HWJJJSoIMaAT58kAIe3tZ+h4upSUsUVPZOX/GrD8WnxqrjaC3oum9+CDy57IkKEpLza1tDRQ/THEz6KdmCh0JI0F95uVwHv6ST+1l9eQPHtA7l+Wbx20DikESozglidZX2PdJSn/gFQr1nFU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1752062006; c=relaxed/simple; bh=gPZ3HY8Z/Qf3HBXo6MxOplLv9ot5uaDonjEvcOp6z+A=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=ftAE2p+2Cz9jc8ZqFUryhCdPfxJBECJLhPW7ynlea1Y5TpUSKCq0RybxELczmV68TzdErj/zMl6wo72MfHcwDf09UJNQGiLAEUFyS8yVcQ+zbz/mLUJCf/hpz3DdBRELrZRlf/6qsHfPcdcB9BsV+TNRpro9bSaNU0im8J123e8= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=BxtkvIHP; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="BxtkvIHP" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 9D569C4CEEF; Wed, 9 Jul 2025 11:53:21 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1752062005; bh=gPZ3HY8Z/Qf3HBXo6MxOplLv9ot5uaDonjEvcOp6z+A=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=BxtkvIHPPN/JEn87h24VP6VXtIxaoK545KYc294WK/rroBBbzEUK4YcJL8+kT9Q9T V9UAVK+Iux4rOX9KXe+A6zESnxlm/+QU7TiSVAJGvCioIR6vdamXB/sioGErhMD9gh jzVAdNhWKaHaJlOCo4nTckcBERTkt13mcRtVa0x3SR8/Ns/N6vHQ+9FcEI8flSS5BE Si6SycY6iDZfRvxhZergTNDuWpuQFNobmykYaUi8LHFv6kfINWSVvTNZdDSzRbynCE KtKNbTG/fyxGWeQDStvzuHUCL2riRSIh/oG81kmxvCKOUySdNslpQycEuUtI3jw9k/ fv8UqcOl2A9kA== From: Philipp Stanner To: Lyude Paul , Danilo Krummrich , David Airlie , Simona Vetter , Matthew Brost , Philipp Stanner , =?UTF-8?q?Christian=20K=C3=B6nig?= , Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , Sumit Semwal , Tvrtko Ursulin , Pierre-Eric Pelloux-Prayer Cc: dri-devel@lists.freedesktop.org, nouveau@lists.freedesktop.org, linux-kernel@vger.kernel.org, linux-media@vger.kernel.org, =?UTF-8?q?Ma=C3=ADra=20Canal?= Subject: [PATCH v3 1/7] drm/sched: Avoid memory leaks with cancel_job() callback Date: Wed, 9 Jul 2025 13:52:51 +0200 Message-ID: <20250709115257.106370-3-phasta@kernel.org> X-Mailer: git-send-email 2.49.0 In-Reply-To: <20250709115257.106370-2-phasta@kernel.org> References: <20250709115257.106370-2-phasta@kernel.org> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Since its inception, the GPU scheduler can leak memory if the driver calls drm_sched_fini() while there are still jobs in flight. The simplest way to solve this in a backwards compatible manner is by adding a new callback, drm_sched_backend_ops.cancel_job(), which instructs the driver to signal the hardware fence associated with the job. Afterwards, the scheduler can safely use the established free_job() callback for freeing the job. Implement the new backend_ops callback cancel_job(). Suggested-by: Tvrtko Ursulin Link: https://lore.kernel.org/dri-devel/20250418113211.69956-1-tvrtko.ursul= in@igalia.com/ Signed-off-by: Philipp Stanner Reviewed-by: Ma=C3=ADra Canal --- drivers/gpu/drm/scheduler/sched_main.c | 34 ++++++++++++++++---------- include/drm/gpu_scheduler.h | 18 ++++++++++++++ 2 files changed, 39 insertions(+), 13 deletions(-) diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/sched= uler/sched_main.c index 81ad40d9582b..a971f0c9e6e0 100644 --- a/drivers/gpu/drm/scheduler/sched_main.c +++ b/drivers/gpu/drm/scheduler/sched_main.c @@ -1352,6 +1352,18 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, = const struct drm_sched_init_ } EXPORT_SYMBOL(drm_sched_init); =20 +static void drm_sched_cancel_remaining_jobs(struct drm_gpu_scheduler *sche= d) +{ + struct drm_sched_job *job, *tmp; + + /* All other accessors are stopped. No locking necessary. */ + list_for_each_entry_safe_reverse(job, tmp, &sched->pending_list, list) { + sched->ops->cancel_job(job); + list_del(&job->list); + sched->ops->free_job(job); + } +} + /** * drm_sched_fini - Destroy a gpu scheduler * @@ -1359,19 +1371,11 @@ EXPORT_SYMBOL(drm_sched_init); * * Tears down and cleans up the scheduler. * - * This stops submission of new jobs to the hardware through - * drm_sched_backend_ops.run_job(). Consequently, drm_sched_backend_ops.fr= ee_job() - * will not be called for all jobs still in drm_gpu_scheduler.pending_list. - * There is no solution for this currently. Thus, it is up to the driver t= o make - * sure that: - * - * a) drm_sched_fini() is only called after for all submitted jobs - * drm_sched_backend_ops.free_job() has been called or that - * b) the jobs for which drm_sched_backend_ops.free_job() has not been ca= lled - * after drm_sched_fini() ran are freed manually. - * - * FIXME: Take care of the above problem and prevent this function from le= aking - * the jobs in drm_gpu_scheduler.pending_list under any circumstances. + * This stops submission of new jobs to the hardware through &struct + * drm_sched_backend_ops.run_job. If &struct drm_sched_backend_ops.cancel_= job + * is implemented, all jobs will be canceled through it and afterwards cle= aned + * up through &struct drm_sched_backend_ops.free_job. If cancel_job is not + * implemented, memory could leak. */ void drm_sched_fini(struct drm_gpu_scheduler *sched) { @@ -1401,6 +1405,10 @@ void drm_sched_fini(struct drm_gpu_scheduler *sched) /* Confirm no work left behind accessing device structures */ cancel_delayed_work_sync(&sched->work_tdr); =20 + /* Avoid memory leaks if supported by the driver. */ + if (sched->ops->cancel_job) + drm_sched_cancel_remaining_jobs(sched); + if (sched->own_submit_wq) destroy_workqueue(sched->submit_wq); sched->ready =3D false; diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h index e62a7214e052..190844370f48 100644 --- a/include/drm/gpu_scheduler.h +++ b/include/drm/gpu_scheduler.h @@ -512,6 +512,24 @@ struct drm_sched_backend_ops { * and it's time to clean it up. */ void (*free_job)(struct drm_sched_job *sched_job); + + /** + * @cancel_job: Used by the scheduler to guarantee remaining jobs' fences + * get signaled in drm_sched_fini(). + * + * Used by the scheduler to cancel all jobs that have not been executed + * with &struct drm_sched_backend_ops.run_job by the time + * drm_sched_fini() gets invoked. + * + * Drivers need to signal the passed job's hardware fence with an + * appropriate error code (e.g., -ECANCELED) in this callback. They + * must not free the job. + * + * The scheduler will only call this callback once it stopped calling + * all other callbacks forever, with the exception of &struct + * drm_sched_backend_ops.free_job. + */ + void (*cancel_job)(struct drm_sched_job *sched_job); }; =20 /** --=20 2.49.0